Using A Disentangled Neural Network to Objectively Assess the Outcomes of Midfacial Surgery in Syndromic Craniosynostosis.

BACKGROUND: Advancements in artificial intelligence and the development of shape models that quantify normal head shape and facial morphology provide frameworks by which the outcomes of craniofacial surgery can be compared. In this work, we will demonstrate the use of the Swap Disentangled Variational Autoencoder (SD-VAE) to objectively assess changes following midfacial surgery. MATERIALS AND METHODS: Our model is trained on a dataset of 1405 3D meshes of healthy and syndromic patients which was augmented using a technique based on spectral interpolation. Patients with a diagnosis of Apert and Crouzon syndrome who had undergone sub- or trans-cranial midfacial procedures utilising rigid external distraction were then interpreted using this model as the point of comparison. RESULTS: A total of 56 patients met our inclusion criteria, 20 with Apert and 36 with Crouzon syndrome. By using linear discriminant analysis to project the high-dimensional vectors derived by SD-VAE onto a 2D space, the shape properties of Apert and Crouzon syndrome can be visualised in relation to the healthy population. In this way, we are able to show how surgery elicits global shape changes in each patient. To assess the regional movements achieved during surgery, we use a novel metric derived from the Malahanobis distance to quantify movements through the latent space. CONCLUSION: Objective outcome evaluation, which encourages in-depth analysis and enhances decision making, is essential for the progression of surgical practice. We have demonstrated how artificial intelligence has the ability to improve our understanding of surgery and its effect on craniofacial morphology.

AI-based diagnosis and phenotype - Genotype correlations in syndromic craniosynostoses.

Apert (AS), Crouzon (CS), Muenke (MS), Pfeiffer (PS), and Saethre Chotzen (SCS) are among the most frequently diagnosed syndromic craniosynostoses. The aims of this study were (1) to train an innovative model using artificial intelligence (AI)-based methods on two-dimensional facial frontal, lateral, and external ear photographs to assist diagnosis for syndromic craniosynostoses vs controls, and (2) to screen for genotype/phenotype correlations in AS, CS, and PS. We included retrospectively and prospectively, from 1979 to 2023, all frontal and lateral pictures of patients genetically diagnosed with AS, CS, MS, PS and SCS syndromes. After a deep learning-based preprocessing, we extracted geometric and textural features and used XGboost (eXtreme Gradient Boosting) to classify patients. The model was tested on an independent international validation set of genetically confirmed patients and non-syndromic controls. Between 1979 and 2023, we included 2228 frontal and lateral facial photographs corresponding to 541 patients. In all, 70.2% [0.593-0.797] (p < 0.001) of patients in the validation set were correctly diagnosed. Genotypes linked to a splice donor site of FGFR2 in Crouzon-Pfeiffer syndrome (CPS) caused a milder phenotype in CPS. Here we report a new method for the automatic detection of syndromic craniosynostoses using AI.

Latent disentanglement in mesh variational autoencoders improves the diagnosis of craniofacial syndromes and aids surgical planning.

BACKGROUND AND OBJECTIVE: The use of deep learning to undertake shape analysis of the complexities of the human head holds great promise. However, there have traditionally been a number of barriers to accurate modelling, especially when operating on both a global and local level. METHODS: In this work, we will discuss the application of the Swap Disentangled Variational Autoencoder (SD-VAE) with relevance to Crouzon, Apert and Muenke syndromes. The model is trained on a dataset of 3D meshes of healthy and syndromic patients which was increased in size with a novel data augmentation technique based on spectral interpolation. Thanks to its semantically meaningful and disentangled latent representation, SD-VAE is used to analyse and generate head shapes while considering the influence of different anatomical sub-units. RESULTS: Although syndrome classification is performed on the entire mesh, it is also possible, for the first time, to analyse the influence of each region of the head on the syndromic phenotype. By manipulating specific parameters of the generative model, and producing procedure-specific new shapes, it is also possible to approximate the outcome of a range of craniofacial surgical procedures. CONCLUSION: This work opens new avenues to advance diagnosis, aids surgical planning and allows for the objective evaluation of surgical outcomes. Our code is available at github.com/simofoti/CraniofacialSD-VAE.

Comparison of Internal and External Distraction in Frontofacial Monobloc Advancement: A Three-Dimensional Quantification.

BACKGROUND: Crouzon syndrome is characterized by complex craniosynostosis and midfacial hypoplasia. Where frontofacial monobloc advancement (FFMBA) is indicated, the method of distraction used to achieve advancement holds an element of equipoise. This two-center retrospective cohort study quantifies the movements produced by internal or external distraction methods used for FFMBA. Using shape analysis, this study evaluates whether the different distraction forces cause plastic deformity of the frontofacial segment, producing distinct morphologic outcomes. METHODS: Patients with Crouzon syndrome who underwent FFMBA with internal distraction [Hôpital Necker-Enfants Malades (Paris, France)] or external distraction [Great Ormond Street Hospital for Children (London, United Kingdom)] were compared. Digital Imaging and Communications in Medicine files of preoperative and postoperative computed tomographic scans were converted to three-dimensional bone meshes and skeletal movements were assessed using nonrigid iterative closest point registration. Displacements were visualized using color maps and statistical analysis of the vectors was undertaken. RESULTS: Fifty-one patients met the strict inclusion criteria. Twenty-five underwent FFMBA with external distraction and 26 with internal distraction. External distraction provides a preferential midfacial advancement, whereas internal distractors produce a more positive movement at the lateral orbital rim. This confers good orbital protection but does not advance the central midface to the same extent. Vector analysis confirmed this to be statistically significant ( P < 0.01). CONCLUSIONS: Morphologic changes resulting from monobloc surgery differ depending on the distraction technique used. Although the relative merits of internal and external distraction still stand, it may be that external distraction is more suited to addressing the midfacial biconcavity seen in syndromic craniosynostosis. CLINICAL QUESTION/LEVEL OF EVIDENCE: Therapeutic, III.

Maxillary Changes Following Facial Bipartition - A Three-Dimensional Quantification.

INTRODUCTION: Children with Apert syndrome have hypertelorism and midfacial hypoplasia, which can be treated with facial bipartition (FB), often aided by rigid external distraction. The technique involves a midline osteotomy that lateralizes the maxillary segments, resulting in posterior cross-bites and midline diastema. Varying degrees of spontaneous realignment of the dental arches occurs postoperatively. This study aims to quantify these movements and assess whether they occur as part of a wider skeletal relapse or as dental compensation. METHODS: Patients who underwent FB and had high quality computed tomography scans at the preoperative stage, immediately postsurgery, and later postoperatively were reviewed. DICOM files were converted to three-dimensional bone meshes and anatomical point-to-point displacements were quantified using nonrigid iterative closest point registration. Displacements were visualized using arrow maps, thereby providing an overview of the movements of the facial skeleton and dentition. RESULTS: Five patients with Apert syndrome were included. In all cases, the arrow maps demonstrated initial significant anterior movement of the frontofacial segment coupled with medial rotation of the orbits and transverse divergence of the maxillary arches. The bony position following initial surgery was shown to be largely stable, with primary dentoalveolar relapse correcting the dental alignment. CONCLUSIONS: This study showed that spontaneous dental compensation occurs following FB without compromising the surgical result. It may be appropriate to delay active orthodontic for 6-months postoperatively until completion of this early compensatory phase.

Facing Africa: Describing Noma in Ethiopia.

Noma affects the most marginalized communities in the world, beginning as oral ulceration and rapidly progressing to orofacial gangrene. With a mortality rate estimated to be as high as 90% and with very few able to access treatment in its active phase, very little is understood about the disease. This retrospective review of patients treated by Facing Africa for deformity and functional impairment secondary to noma between May 2015 and 2019 highlights some of the difficulties encountered by those afflicted. Eighty new patients with historical noma defects were identified and were seen over the course of nine surgical missions, with notes providing valuable geographical, socioeconomic, and psychosocial information. The mean self-reported age of onset was 5 years and 8 months, with a median time of 18 years from onset to accessing treatment. Before intervention, 65% covered their face in public, 59% reported difficulty eating, 81% were unhappy with their appearance, and 71% experienced bullying. We aimed at emphasizing the significant burden, both psychologically and physically of noma, demonstrating the disparity between recent decades of progress in the well-being of Ethiopians in general and the access to health care and mental health support for some of those most in need.